Today, large sets of servers may be configured to host various websites, web applications, and/or web services. For example, a set of servers, such as a hundred servers, may be configured to process user requests from thousands of users consuming an email web service. If an issue arises with a user attempting to access the email web service, then limited troubleshooting may be performed server-side and/or client-side. In one example, troubleshooting the issue client-side may call for the user to install tracing tools configured to trace network traffic between the client machine and a server processing user requests from the client machine, which may be burdensome to the user. In another example, tracing may be enabled server-side on all servers because it may be unknown which server may process users requests from the client machine, which may result in expensive tracing data generated for all user requests from all users. In another example, a small number of servers may be isolated from the set of servers, which may be a laborious process and/or necessitate near-real-time coordination between a troubleshooting administrator and the user. The isolated set of servers may be dedicated to processing and/or tracing user requests from the user in order to troubleshoot the issue. In this way, the tracing data may be utilized in troubleshooting and/or debugging issues.